<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Task-Specification on 兜忆轩 | Cedric's Compiler Notes</title><link>https://douyixuan.github.io/tags/task-specification/</link><description>Recent content in Task-Specification on 兜忆轩 | Cedric's Compiler Notes</description><image><title>兜忆轩 | Cedric's Compiler Notes</title><url>https://douyixuan.github.io/images/papermod-cover.png</url><link>https://douyixuan.github.io/images/papermod-cover.png</link></image><generator>Hugo -- 0.146.0</generator><language>en-us</language><lastBuildDate>Fri, 28 Aug 2026 08:00:00 +0800</lastBuildDate><atom:link href="https://douyixuan.github.io/tags/task-specification/index.xml" rel="self" type="application/rss+xml"/><item><title>AI Native Digest 2026-08-28</title><link>https://douyixuan.github.io/posts/ai-native/2026-08-28/</link><pubDate>Fri, 28 Aug 2026 08:00:00 +0800</pubDate><guid>https://douyixuan.github.io/posts/ai-native/2026-08-28/</guid><description>今天两篇都在谈 AI software factory 的单位经济学：Uber 已经把 70%+ PR、3,600+ skills 和 30K+ daily skill executions 推到生产规模，并把 agent 成本拆成可优化的工程指标；一篇新 arXiv 研究则量化了 task specification 对 coding-agent token spend 的影响，说明更完整、更具体的任务说明往往能直接减少 Agent 自己的搜索与推理成本。</description></item></channel></rss>